公开(公告)号：US20220329936A1

公开(公告)日：2022-10-13

申请号：US17640305

申请日：2019-09-09

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kazunori KOBAYASHI ,  Hiroaki ITO ,  Shin MURATA

IPC分类号： H04R3/00

摘要： Audibility of an outside sound needed for a driver inside an automobile to apprehend a danger and obtain a grasp of a situation necessary for driving is improved. A sound collection and emission apparatus (10) emits, on the basis of an outside acoustic signal which emanates from a sound source outside an automobile (90) and arrives at the automobile (90), an inside acoustic signal which is an acoustic signal derived from the outside acoustic signal to inside the automobile (90). A sound collection unit (M1) collects the outside acoustic signal. A sound emission unit (S1) emits the inside acoustic signal. A danger sound detection unit (11) determines whether the outside acoustic signal has a feature representing a danger defined in advance. A control unit (12) performs control that emits the inside acoustic signal from the sound emission unit (S1) such that a driver of the automobile (90) is capable of perceiving the danger if the outside acoustic signal is determined to represent the danger.

公开(公告)号：US20200209842A1

公开(公告)日：2020-07-02

申请号：US16643916

申请日：2018-08-24

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yuma KOIZUMI ,  Yuta KAWACHI ,  Noboru HARADA ,  Shoichiro SAITO ,  Akira NAKAGAWA ,  Shin MURATA

IPC分类号： G05B23/02 ,  G06N20/00

摘要： Accuracy of unsupervised anomalous sound detection is improved using a small number of pieces of anomalous sound data. A threshold deciding part (13) calculates an anomaly score for each of a plurality of pieces of anomalous sound data, using a normal model learned with normal sound data and an anomaly model expressing the pieces of anomalous sound data, and decides a minimum value among the anomaly scores as a threshold. A weight updating part (14) updates, using a plurality of pieces of normal sound data, the pieces of anomalous sound data and the threshold, weights of the anomaly model so that all the pieces of anomalous sound data are judged as anomalous, and probability of the pieces of normal sound data being judged as anomalous is minimized.

公开(公告)号：US20220327379A1

公开(公告)日：2022-10-13

申请号：US17639330

申请日：2019-09-02

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yuma KOIZUMI ,  Shin MURATA ,  Ryotaro SATO

IPC分类号： G06N3/08

摘要： There is provided a neural network learning technique for learning a parameter of a probability density function representing the distribution of data with high accuracy using an autoencoder. A neural network learning apparatus, wherein θ is a parameter of a probability density function qθ(x) representing distribution of data x, and Mθ is a neural network that is an autoencoder that learns the parameter θ, the neural network learning apparatus including: a neural network calculation unit that calculates an output value Mθ(xn) of the neural network from learning data xn using the parameter θ for n=1, . . . , N; a cost function calculation unit that calculates an evaluation value of a cost function L using the learning data xn (1≤n≤N) and the output value Mθ(xn) (1≤n≤N); and a parameter update unit that updates the parameter θ using the evaluation value, wherein the cost function L is defined by an expression using a normalization constant Zθ of a Boltzmann distribution defined based on a reconstruction error Eθ(x)=∥x−Mθ(x)∥22 of the data x.

公开(公告)号：US20220250540A1

公开(公告)日：2022-08-11

申请号：US17621727

申请日：2019-06-26

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kazunori KOBAYASHI ,  Hiroaki ITO ,  Shin MURATA

IPC分类号： B60Q5/00 ,  B60W50/14 ,  G08G1/16

摘要： A danger detection system for detecting a danger on a road, the danger that changes over time and may not be a danger due to the change over time, the danger detection system including a vibration acquisition unit 11 configured to acquire an acoustic signal collected by a microphone mounted on a vehicle when the vehicle passes through a region on a predetermined road, a danger determination unit 12 configured to determine whether the acoustic signal is derived from the danger, a danger location detection unit 2 configured to determine whether the danger is present in the region on the predetermined road based on the number of times an acoustic signal acquired from each of the plurality of vehicles is determined to be derived from the danger, and a danger information notification unit 3 configured to notify a communication device that can pass through the region on the predetermined road of danger information that is information about a region on a road determined to have a danger.

公开(公告)号：US20220219703A1

公开(公告)日：2022-07-14

申请号：US17610144

申请日：2019-05-16

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kazunori KOBAYASHI ,  Hiroaki ITO ,  Shin MURATA

IPC分类号： B60W40/06 ,  G01H11/06 ,  H04L27/26

摘要： An abnormality detection device 1 includes: a vibration acquisition unit 11 configured to acquire an acoustic signal generated during passage of a vehicle on a road; a frequency domain conversion unit 12 configured to convert the acquired acoustic signal into a frequency domain signal; an unexpectedness determination unit 13 configured to determine whether there is unexpectedness at each predetermined frequency using the frequency domain signal; and an abnormality determination unit 14 configured to determine whether there is an abnormality in the road based on the number of frequencies at which it is determined that there is the unexpectedness.

公开(公告)号：US20230086628A1

公开(公告)日：2023-03-23

申请号：US17798849

申请日：2020-02-12

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yuma KOIZUMI ,  Shoichiro SAITO ,  Hisashi UEMATSU ,  Shin MURATA

IPC分类号： G06K9/62 ,  G06N3/04

摘要： Provided is an abnormal data generation device capable of generating highly accurate abnormal data. The abnormal data generation device includes an abnormal data generation unit for generating pseudo generated data of abnormal data that has, in the same latent space, a normal distribution as a normal data generation model and an abnormal distribution expressed as a complementary set of the normal distribution and that is optimized such that pseudo generated data cannot be discriminated from observed actual abnormal data by a latent variable sampled from the abnormal distribution.

公开(公告)号：US20210326728A1

公开(公告)日：2021-10-21

申请号：US17260956

申请日：2019-07-01

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yuta KAWACHI ,  Yuma KOIZUMI ,  Noboru HARADA ,  Shin MURATA

IPC分类号： G06N5/04 ,  G06N20/00

摘要： A possible region of encoding results of anomalous samples is limited. An encoder storage unit 14 stores an encoder for projecting an input feature value into a latent space in which the latent space is a closed manifold, a normal distribution obtained by learning normal data and an anomalous distribution obtained by learning anomalous data are held on the manifold, and a decoder for reconstructing the output of the encoder. An encoding unit 15 obtains a reconstruction result output by the decoder when a feature value of target data is input to the encoder. An anomaly score calculation unit 16 calculates an anomaly score of the target data based on distances between the reconstruction result and the normal distribution and distances between the reconstruction result and the anomalous distribution.

公开(公告)号：US20240152133A1

公开(公告)日：2024-05-09

申请号：US17769295

申请日：2019-10-16

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Shin MURATA ,  Yuma KOIZUMI ,  Noboru HARADA ,  Shoichiro SAITO

IPC分类号： G06F11/07

CPC分类号： G06F11/076 ,  G06F11/0706

摘要： A threshold acquisition apparatus acquires a threshold for determining whether an anomaly score acquired from a target sound is normal or anomalous. The threshold acquisition apparatus includes: an allowable number setting unit that sets an allowable number of times such that the number of anomaly scores determined to be anomalous included in a set of anomaly scores per predetermined section length, which is a part of time-series acoustic signals that do not include an anomalous sound, does not exceed the allowable number of times; and a threshold estimation unit that estimates a threshold candidate such that the number of sections determined to be anomalous per predetermined section length, which is a part of time-series acoustic signals, satisfies a predetermined criterion by using the allowable number of times, and acquires the threshold candidate as the threshold.

公开(公告)号：US20230162755A1

公开(公告)日：2023-05-25

申请号：US17916963

申请日：2020-04-08

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kazunori KOBAYASHI ,  Shin MURATA

IPC分类号： G10L25/51 ,  G10L25/21 ,  G10L25/18 ,  G10L25/90 ,  G10L25/06

CPC分类号： G10L25/51 ,  G10L25/21 ,  G10L25/18 ,  G10L25/90 ,  G10L25/06

摘要： In order to detect the object left-behind of an infant in an automobile without installing a dedicated sensor, a microphone (M1) installed in the automobile picks up an acoustic signal. An autocorrelation unit (11) determines an autocorrelation function from the acoustic signal. A peak detection unit (12) detects the time of a peak of the autocorrelation value as a pitch period. An inverse calculation unit (13) calculates the inverse of a peak period as a pitch frequency. A pitch determination unit (21) determines whether or not the pitch frequency is included in a predetermined frequency band.

公开(公告)号：US20220289131A1

公开(公告)日：2022-09-15

申请号：US17632527

申请日：2019-08-05

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Hiroaki ITO ,  Shin MURATA

IPC分类号： B60R21/0132 ,  B60W30/08

摘要： Provided is an anomaly detection device capable of acquiring a degree of anomaly required for anomaly detection even when a feature pattern of time series data changes over time. The anomaly detection device of the present invention is a device that detects the degree of anomaly in the time series data. The anomaly detection device includes a first acquisition unit, a prediction unit, and an anomaly degree acquisition unit. The first acquisition unit acquires, from a first section that is a partial section of the time series data, a dynamic feature pattern of the first section. The prediction unit predicts data of a second section that is a partial section of the time series data later than the first section by using the feature pattern to obtain predicted second section data. The anomaly degree acquisition unit acquires the degree of anomaly based on a difference between the predicted second section data and actual data of the second section in the time series data.